[PATCH 10/20] perf tools: Reconstruct hw cache event with modifiers from perf_event_attr
From: Arnaldo Carvalho de Melo
Date: Tue Jun 19 2012 - 14:05:03 EST
From: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
[root@sandy ~]# perf record -a -e dTLB-load-misses:u usleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.486 MB perf.data (~21216 samples) ]
Before:
[root@sandy ~]# perf evlist
dTLB-load-misses
[root@sandy ~]#
After:
[root@sandy ~]# perf evlist
dTLB-load-misses:u
[root@sandy ~]#
Ditto for other tools.
Cc: David Ahern <dsahern@xxxxxxxxx>
Cc: Frederic Weisbecker <fweisbec@xxxxxxxxx>
Cc: Jiri Olsa <jolsa@xxxxxxxxxx>
Cc: Mike Galbraith <efault@xxxxxx>
Cc: Namhyung Kim <namhyung@xxxxxxxxx>
Cc: Paul Mackerras <paulus@xxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Stephane Eranian <eranian@xxxxxxxxxx>
Link: http://lkml.kernel.org/n/tip-7x1b0e6jthkr93lfjzsuakk5@xxxxxxxxxxxxxx
Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
---
tools/perf/util/evsel.c | 104 ++++++++++++++++++++++++++++++++++++
tools/perf/util/evsel.h | 16 +++++-
tools/perf/util/parse-events.c | 115 +++++----------------------------------
3 files changed, 134 insertions(+), 101 deletions(-)
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index dab8938..47f1fe2 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -128,6 +128,105 @@ static int perf_evsel__hw_name(struct perf_evsel *evsel, char *bf, size_t size)
return r + perf_evsel__add_modifiers(evsel, bf + r, size - r);
}
+const char *perf_evsel__hw_cache[PERF_COUNT_HW_CACHE_MAX]
+ [PERF_EVSEL__MAX_ALIASES] = {
+ { "L1-dcache", "l1-d", "l1d", "L1-data", },
+ { "L1-icache", "l1-i", "l1i", "L1-instruction", },
+ { "LLC", "L2", },
+ { "dTLB", "d-tlb", "Data-TLB", },
+ { "iTLB", "i-tlb", "Instruction-TLB", },
+ { "branch", "branches", "bpu", "btb", "bpc", },
+ { "node", },
+};
+
+const char *perf_evsel__hw_cache_op[PERF_COUNT_HW_CACHE_OP_MAX]
+ [PERF_EVSEL__MAX_ALIASES] = {
+ { "load", "loads", "read", },
+ { "store", "stores", "write", },
+ { "prefetch", "prefetches", "speculative-read", "speculative-load", },
+};
+
+const char *perf_evsel__hw_cache_result[PERF_COUNT_HW_CACHE_RESULT_MAX]
+ [PERF_EVSEL__MAX_ALIASES] = {
+ { "refs", "Reference", "ops", "access", },
+ { "misses", "miss", },
+};
+
+#define C(x) PERF_COUNT_HW_CACHE_##x
+#define CACHE_READ (1 << C(OP_READ))
+#define CACHE_WRITE (1 << C(OP_WRITE))
+#define CACHE_PREFETCH (1 << C(OP_PREFETCH))
+#define COP(x) (1 << x)
+
+/*
+ * cache operartion stat
+ * L1I : Read and prefetch only
+ * ITLB and BPU : Read-only
+ */
+static unsigned long perf_evsel__hw_cache_stat[C(MAX)] = {
+ [C(L1D)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+ [C(L1I)] = (CACHE_READ | CACHE_PREFETCH),
+ [C(LL)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+ [C(DTLB)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+ [C(ITLB)] = (CACHE_READ),
+ [C(BPU)] = (CACHE_READ),
+ [C(NODE)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
+};
+
+bool perf_evsel__is_cache_op_valid(u8 type, u8 op)
+{
+ if (perf_evsel__hw_cache_stat[type] & COP(op))
+ return true; /* valid */
+ else
+ return false; /* invalid */
+}
+
+int __perf_evsel__hw_cache_type_op_res_name(u8 type, u8 op, u8 result,
+ char *bf, size_t size)
+{
+ if (result) {
+ return scnprintf(bf, size, "%s-%s-%s", perf_evsel__hw_cache[type][0],
+ perf_evsel__hw_cache_op[op][0],
+ perf_evsel__hw_cache_result[result][0]);
+ }
+
+ return scnprintf(bf, size, "%s-%s", perf_evsel__hw_cache[type][0],
+ perf_evsel__hw_cache_op[op][1]);
+}
+
+int __perf_evsel__hw_cache_name(u64 config, char *bf, size_t size)
+{
+ u8 op, result, type = (config >> 0) & 0xff;
+ const char *err = "unknown-ext-hardware-cache-type";
+
+ if (type > PERF_COUNT_HW_CACHE_MAX)
+ goto out_err;
+
+ op = (config >> 8) & 0xff;
+ err = "unknown-ext-hardware-cache-op";
+ if (op > PERF_COUNT_HW_CACHE_OP_MAX)
+ goto out_err;
+
+ result = (config >> 16) & 0xff;
+ err = "unknown-ext-hardware-cache-result";
+ if (result > PERF_COUNT_HW_CACHE_RESULT_MAX)
+ goto out_err;
+
+ err = "invalid-cache";
+ if (!perf_evsel__is_cache_op_valid(type, op))
+ goto out_err;
+
+ return __perf_evsel__hw_cache_type_op_res_name(type, op, result, bf, size);
+out_err:
+ return scnprintf(bf, size, "%s", err);
+}
+
+static int perf_evsel__hw_cache_name(struct perf_evsel *evsel, char *bf, size_t size)
+{
+ int ret = __perf_evsel__hw_cache_name(evsel->attr.config, bf, size);
+ return ret + perf_evsel__add_modifiers(evsel, bf + ret, size - ret);
+}
+
int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size)
{
int ret;
@@ -140,6 +239,11 @@ int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size)
case PERF_TYPE_HARDWARE:
ret = perf_evsel__hw_name(evsel, bf, size);
break;
+
+ case PERF_TYPE_HW_CACHE:
+ ret = perf_evsel__hw_cache_name(evsel, bf, size);
+ break;
+
default:
/*
* FIXME
diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
index 4ba8b56..5bf946a 100644
--- a/tools/perf/util/evsel.h
+++ b/tools/perf/util/evsel.h
@@ -83,7 +83,21 @@ void perf_evsel__config(struct perf_evsel *evsel,
struct perf_record_opts *opts,
struct perf_evsel *first);
-const char* __perf_evsel__hw_name(u64 config);
+bool perf_evsel__is_cache_op_valid(u8 type, u8 op);
+
+#define PERF_EVSEL__MAX_ALIASES 8
+
+extern const char *perf_evsel__hw_cache[PERF_COUNT_HW_CACHE_MAX]
+ [PERF_EVSEL__MAX_ALIASES];
+extern const char *perf_evsel__hw_cache_op[PERF_COUNT_HW_CACHE_OP_MAX]
+ [PERF_EVSEL__MAX_ALIASES];
+const char *perf_evsel__hw_cache_result[PERF_COUNT_HW_CACHE_RESULT_MAX]
+ [PERF_EVSEL__MAX_ALIASES];
+int __perf_evsel__hw_cache_type_op_res_name(u8 type, u8 op, u8 result,
+ char *bf, size_t size);
+int __perf_evsel__hw_cache_name(u64 config, char *bf, size_t size);
+
+const char *__perf_evsel__hw_name(u64 config);
int perf_evsel__name(struct perf_evsel *evsel, char *bf, size_t size);
int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads);
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 05dbc8b..c8f8cf4 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -74,51 +74,6 @@ static const char *sw_event_names[PERF_COUNT_SW_MAX] = {
"emulation-faults",
};
-#define MAX_ALIASES 8
-
-static const char *hw_cache[PERF_COUNT_HW_CACHE_MAX][MAX_ALIASES] = {
- { "L1-dcache", "l1-d", "l1d", "L1-data", },
- { "L1-icache", "l1-i", "l1i", "L1-instruction", },
- { "LLC", "L2", },
- { "dTLB", "d-tlb", "Data-TLB", },
- { "iTLB", "i-tlb", "Instruction-TLB", },
- { "branch", "branches", "bpu", "btb", "bpc", },
- { "node", },
-};
-
-static const char *hw_cache_op[PERF_COUNT_HW_CACHE_OP_MAX][MAX_ALIASES] = {
- { "load", "loads", "read", },
- { "store", "stores", "write", },
- { "prefetch", "prefetches", "speculative-read", "speculative-load", },
-};
-
-static const char *hw_cache_result[PERF_COUNT_HW_CACHE_RESULT_MAX]
- [MAX_ALIASES] = {
- { "refs", "Reference", "ops", "access", },
- { "misses", "miss", },
-};
-
-#define C(x) PERF_COUNT_HW_CACHE_##x
-#define CACHE_READ (1 << C(OP_READ))
-#define CACHE_WRITE (1 << C(OP_WRITE))
-#define CACHE_PREFETCH (1 << C(OP_PREFETCH))
-#define COP(x) (1 << x)
-
-/*
- * cache operartion stat
- * L1I : Read and prefetch only
- * ITLB and BPU : Read-only
- */
-static unsigned long hw_cache_stat[C(MAX)] = {
- [C(L1D)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
- [C(L1I)] = (CACHE_READ | CACHE_PREFETCH),
- [C(LL)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
- [C(DTLB)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
- [C(ITLB)] = (CACHE_READ),
- [C(BPU)] = (CACHE_READ),
- [C(NODE)] = (CACHE_READ | CACHE_WRITE | CACHE_PREFETCH),
-};
-
#define for_each_subsystem(sys_dir, sys_dirent, sys_next) \
while (!readdir_r(sys_dir, &sys_dirent, &sys_next) && sys_next) \
if (sys_dirent.d_type == DT_DIR && \
@@ -236,30 +191,6 @@ static const char *tracepoint_id_to_name(u64 config)
return buf;
}
-static int is_cache_op_valid(u8 cache_type, u8 cache_op)
-{
- if (hw_cache_stat[cache_type] & COP(cache_op))
- return 1; /* valid */
- else
- return 0; /* invalid */
-}
-
-static char *event_cache_name(u8 cache_type, u8 cache_op, u8 cache_result)
-{
- static char name[50];
-
- if (cache_result) {
- sprintf(name, "%s-%s-%s", hw_cache[cache_type][0],
- hw_cache_op[cache_op][0],
- hw_cache_result[cache_result][0]);
- } else {
- sprintf(name, "%s-%s", hw_cache[cache_type][0],
- hw_cache_op[cache_op][1]);
- }
-
- return name;
-}
-
const char *event_type(int type)
{
switch (type) {
@@ -287,7 +218,7 @@ const char *event_name(struct perf_evsel *evsel)
u64 config = evsel->attr.config;
int type = evsel->attr.type;
- if (type == PERF_TYPE_RAW || type == PERF_TYPE_HARDWARE) {
+ if (type == PERF_TYPE_RAW || type == PERF_TYPE_HARDWARE || type == PERF_TYPE_HW_CACHE) {
/*
* XXX minimal fix, see comment on perf_evsen__name, this static buffer
* will go away together with event_name in the next devel cycle.
@@ -316,26 +247,9 @@ const char *__event_name(int type, u64 config)
case PERF_TYPE_HARDWARE:
return __perf_evsel__hw_name(config);
- case PERF_TYPE_HW_CACHE: {
- u8 cache_type, cache_op, cache_result;
-
- cache_type = (config >> 0) & 0xff;
- if (cache_type > PERF_COUNT_HW_CACHE_MAX)
- return "unknown-ext-hardware-cache-type";
-
- cache_op = (config >> 8) & 0xff;
- if (cache_op > PERF_COUNT_HW_CACHE_OP_MAX)
- return "unknown-ext-hardware-cache-op";
-
- cache_result = (config >> 16) & 0xff;
- if (cache_result > PERF_COUNT_HW_CACHE_RESULT_MAX)
- return "unknown-ext-hardware-cache-result";
-
- if (!is_cache_op_valid(cache_type, cache_op))
- return "invalid-cache";
-
- return event_cache_name(cache_type, cache_op, cache_result);
- }
+ case PERF_TYPE_HW_CACHE:
+ __perf_evsel__hw_cache_name(config, buf, sizeof(buf));
+ return buf;
case PERF_TYPE_SOFTWARE:
if (config < PERF_COUNT_SW_MAX && sw_event_names[config])
@@ -379,13 +293,13 @@ static int add_event(struct list_head **_list, int *idx,
return 0;
}
-static int parse_aliases(char *str, const char *names[][MAX_ALIASES], int size)
+static int parse_aliases(char *str, const char *names[][PERF_EVSEL__MAX_ALIASES], int size)
{
int i, j;
int n, longest = -1;
for (i = 0; i < size; i++) {
- for (j = 0; j < MAX_ALIASES && names[i][j]; j++) {
+ for (j = 0; j < PERF_EVSEL__MAX_ALIASES && names[i][j]; j++) {
n = strlen(names[i][j]);
if (n > longest && !strncasecmp(str, names[i][j], n))
longest = n;
@@ -410,7 +324,7 @@ int parse_events_add_cache(struct list_head **list, int *idx,
* No fallback - if we cannot get a clear cache type
* then bail out:
*/
- cache_type = parse_aliases(type, hw_cache,
+ cache_type = parse_aliases(type, perf_evsel__hw_cache,
PERF_COUNT_HW_CACHE_MAX);
if (cache_type == -1)
return -EINVAL;
@@ -423,18 +337,18 @@ int parse_events_add_cache(struct list_head **list, int *idx,
snprintf(name + n, MAX_NAME_LEN - n, "-%s\n", str);
if (cache_op == -1) {
- cache_op = parse_aliases(str, hw_cache_op,
+ cache_op = parse_aliases(str, perf_evsel__hw_cache_op,
PERF_COUNT_HW_CACHE_OP_MAX);
if (cache_op >= 0) {
- if (!is_cache_op_valid(cache_type, cache_op))
+ if (!perf_evsel__is_cache_op_valid(cache_type, cache_op))
return -EINVAL;
continue;
}
}
if (cache_result == -1) {
- cache_result = parse_aliases(str, hw_cache_result,
- PERF_COUNT_HW_CACHE_RESULT_MAX);
+ cache_result = parse_aliases(str, perf_evsel__hw_cache_result,
+ PERF_COUNT_HW_CACHE_RESULT_MAX);
if (cache_result >= 0)
continue;
}
@@ -970,16 +884,17 @@ void print_events_type(u8 type)
int print_hwcache_events(const char *event_glob)
{
unsigned int type, op, i, printed = 0;
+ char name[64];
for (type = 0; type < PERF_COUNT_HW_CACHE_MAX; type++) {
for (op = 0; op < PERF_COUNT_HW_CACHE_OP_MAX; op++) {
/* skip invalid cache type */
- if (!is_cache_op_valid(type, op))
+ if (!perf_evsel__is_cache_op_valid(type, op))
continue;
for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
- char *name = event_cache_name(type, op, i);
-
+ __perf_evsel__hw_cache_type_op_res_name(type, op, i,
+ name, sizeof(name));
if (event_glob != NULL && !strglobmatch(name, event_glob))
continue;
--
1.7.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/